13:55:54 RRSAgent has joined #sparql 13:55:54 logging to http://www.w3.org/2010/10/05-sparql-irc 13:55:58 does the speaker phone work? 13:56:02 yup 13:56:05 cool 13:56:10 no one else here so hopefully i won't bother anyone :) 13:56:12 pretty well, i hear, from those who have to listen to it 13:56:51 trackbot, start meeting 13:56:53 RRSAgent, make logs world 13:56:53 Zakim has joined #sparql 13:56:55 Zakim, this will be 77277 13:56:55 ok, trackbot; I see SW_(SPARQL)10:00AM scheduled to start in 4 minutes 13:56:56 Meeting: SPARQL Working Group Teleconference 13:56:56 Date: 05 October 2010 13:57:00 zakim, this will be SPARQL 13:57:00 ok, LeeF, I see SW_(SPARQL)10:00AM already started 13:57:10 Agenda: http://www.w3.org/2009/sparql/wiki/Agenda-2010-10-05 13:57:13 Regrets: NickH 13:57:15 SteveH__ has joined #sparql 13:57:23 Chair: LeeF 13:57:29 Scribenick: AxelPolleres 13:57:34 zakim, please call ericp-office 13:57:34 ok, LeeF; the call is being made 13:57:35 +Ericp 13:57:41 zakim, hello 13:57:41 I don't understand 'hello', AndyS 13:57:50 wow, I got a line :-) 13:57:52 + +44.186.528.aaaa 13:57:53 +Sandro 13:57:57 +??P9 13:58:02 Zakim, +44.186.528.aaaa is me 13:58:02 +bglimm; got it 13:58:07 +kasei 13:58:08 Zakim, mute me 13:58:08 bglimm should now be muted 13:58:09 zakim, ??P9 is me 13:58:10 +AndyS; got it 13:58:30 zakim, who is talking 13:58:30 I don't understand 'who is talking', sandro 13:58:30 zakim, who's talking? 13:58:33 zakim, who is talking? 13:58:36 is it gone? 13:58:41 LeeF, listening for 10 seconds I heard sound from the following: ??P1 (0%), AndyS (22%) 13:58:45 + +33.4.92.38.aabb 13:58:51 sandro, listening for 10 seconds I heard sound from the following: ??P1 (4%), AndyS (64%) 13:58:56 + +44.208.439.aacc 13:58:56 -AndyS 13:59:10 chimezie has joined #sparql 13:59:13 Zakim, aacc is Garlik 13:59:13 +Garlik; got it 13:59:15 zakim, dial ivan-voip 13:59:15 ok, ivan; the call is being made 13:59:16 +Ivan 13:59:19 Zakim, aabb is me 13:59:19 +OlivierCorby; got it 13:59:25 Zakim, Garlik is temporarily me 13:59:25 +SteveH__; got it 13:59:25 zakim, who's here? 13:59:26 On the phone I see ??P1, Ericp, bglimm (muted), Sandro, kasei, OlivierCorby, SteveH__, Ivan 13:59:29 On IRC I see chimezie, SteveH__, Zakim, RRSAgent, bglimm, AxelPolleres, NicoM, OlivierCorby, AndyS, LeeF, SteveH_, ivan, karl, iv_an_ru, AlexPassant, pgearon, trackbot, ericP, 13:59:32 ... kasei, sandro 13:59:32 +??P16 13:59:34 MattPerry has joined #sparql 13:59:52 AndyS's line sounds bad to me too 14:00:00 ah, I've got the bad line! 14:00:12 zakim, ??P16 is me 14:00:12 +AndyS; got it 14:00:21 sounds better now 14:00:25 zakim, ??P1 is AxelPolleres 14:00:25 +AxelPolleres; got it 14:00:28 a bit 14:00:30 zakim, who's on the phone? 14:00:30 On the phone I see AxelPolleres, Ericp, bglimm (muted), Sandro, kasei, OlivierCorby, SteveH__, Ivan, AndyS 14:00:30 Zakim, what is the passcode? 14:00:34 the conference code is 77277 (tel:+1.617.761.6200 tel:+33.4.26.46.79.03 tel:+44.203.318.0479), chimezie 14:00:39 echo is back 14:00:42 zakim, ericp temporarily has me 14:00:42 I don't understand 'ericp temporarily has me', LeeF 14:00:44 zakim, ericp temporarily is LeeF 14:00:44 I don't understand 'ericp temporarily is LeeF', LeeF 14:00:52 bglimm, but the cat can chase them around 14:00:52 zakim, ericp is LeeF 14:00:52 +LeeF; got it 14:00:53 + +1.603.897.aadd 14:01:07 +Chimezie_Ogbuji 14:01:08 zakim, aadd is me 14:01:08 +MattPerry; got it 14:01:24 +??P20 14:01:30 Zakim, ??P20 is me 14:01:30 +AlexPassant; got it 14:01:44 yup 14:01:50 scribenick: AxelPolleres 14:01:54 topic: Admin 14:01:57 PROPOSED: Approve minutes at http://www.w3.org/2009/sparql/meeting/2010-09-28 14:01:58 Zakim, mute me 14:01:58 Chimezie_Ogbuji should now be muted 14:01:59 topic: admin 14:02:30 RESOLVED: Approve minutes at http://www.w3.org/2009/sparql/meeting/2010-09-28 14:02:45 Next regular meeting: 2010-10-12 @ 15:00 UK / 10:00 EDT (scribe: SteveH ) 14:02:50 regrets for next week 14:02:52 Regrets for next week, I'm traveling 14:02:52 next meeting usual time next week 14:02:56 and the week after, actually 14:03:16 oops, wait! I will be around next week, only the week after will not be 14:03:17 sorry 14:03:26 subtopic: schedule/extension 14:03:28 Regrets for next two weeks. Will also be travelling 14:03:37 echo is not good 14:03:47 but can understand if concentrate 14:03:49 http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0005.html 14:04:38 plan is to have LC Dec 1, CR Feb 1, PR Apr 15, Rec by June 1 14:04:48 Souri has joined #sparql 14:05:22 feedback from Andy and Steve... Dec 1 might be too tight, rather Dec 15 14:05:23 + +1.603.897.aaee 14:05:55 ... andy had questions about length of LC period length. 14:06:15 zakim, aaee is me 14:06:15 +Souri; got it 14:07:57 LC : Dec 15? 14:08:08 + +1.562.978.aaff 14:08:10 -AxelPolleres 14:08:19 CR : mar 1 14:08:20 LeeF: don't count 2 weeks of holidays 14:08:21 Zakim, aaff is me 14:08:21 +AxelPolleres; got it 14:08:26 can scribe again! 14:08:27 ... CR at March 1 14:08:34 ... 2.5 month CR 14:08:36 PR : May 15 14:08:45 REC: Jul 1 14:09:25 q+ to ask how protocol/tests fit into this 14:09:32 Andy: don't think we need LC period longer, but we need to handle all comments. 14:09:47 Ivan: Christmas vacation closes down half of the world. 14:10:33 LeeF: andy, what you say is that if we receive comments for 6 weeks we need probably another 4 weeks for addressing them 14:10:58 ... that would help, yes. 14:11:21 Andy: LC Dec 15... how much time in advance do we need to review all docs? 14:12:00 LeeF: we need to be approving latest 1st week of Dec, reviews two weeks before that, maybe mid november docs need to be finished 14:12:25 ... 1 1/2 months from now... severe changes after that will need to be handled carefully, documented well 14:12:36 LeeF: Docs need to be basically finished by Nov 15 14:13:01 ... challenge if we go any later is that we would slip another month with the schedule. 14:13:10 ack kasei 14:13:10 kasei, you wanted to ask how protocol/tests fit into this 14:13:31 -OlivierCorby 14:13:33 greg: how does this schedule affect protocol and test suite? 14:13:48 Lee: hope for a sprint on the protocol doc 14:14:00 (Axel: same here for Test Suite) 14:14:19 +OlivierCorby 14:14:23 Lee: hesitant to ask someone else at this point, but we lost an editor. 14:14:42 ... I am less concerned about test suite, enough if it comes together by CR 14:14:51 ... but of course better if earlier. 14:15:08 ... don't want to do a separate schedule for those two, though. 14:15:36 ... pushing out further would mean not finish until August. 14:16:00 ... let's rather try Dec 15 the best we can. 14:16:26 topic: WD publication 14:16:31 ... chairs & team will take care of charter extensions, respectively. 14:16:40 -OlivierCorby 14:17:02 let's start with Query (andy&steve) 14:17:06 Andy: I am done. 14:17:49 Steve: It's fine from my side. 14:17:58 yes 14:17:58 +OlivierCorby 14:18:00 yes, sound is better now 14:18:11 LeeF: two reviewers agreed to publish last week. 14:18:27 PROPOSED: Publish http://www.w3.org/2009/sparql/docs/query-1.1/ as working draft 14:18:36 +1 14:18:36 +1 14:18:37 +1 14:18:37 +1 14:18:51 +1 14:18:55 RESOLVED: Publish http://www.w3.org/2009/sparql/docs/query-1.1/ as working draft 14:19:11 subtopic: update 14:19:39 LeeF: paul/alex did you have a chance to include Axel's/Andy's comments? 14:19:50 Alex: worked in almost all comments 14:20:21 -SteveH__ 14:20:26 ... will be able to address formal model comments this week 14:20:39 +Garlik 14:20:41 Andy: I am fine with publishing "as is" 14:20:47 Zakim, Garlik is temporarily me 14:20:47 +SteveH__; got it 14:21:37 PROPOSED: Publish http://www.w3.org/2009/sparql/docs/update-1.1/ following Axel's OK on the 2nd half of his review 14:21:43 +Philippe 14:22:06 +1 14:22:20 +1 14:22:22 +1 14:22:24 RESOLVED: Publish http://www.w3.org/2009/sparql/docs/update-1.1/ following Axel's OK on the 2nd half of his review 14:22:31 ACTION: Axel to complete update review in the next two days. 14:22:31 Created ACTION-321 - Complete update review in the next two days. [on Axel Polleres - due 2010-10-12]. 14:22:35 Zakim, unmute me 14:22:35 bglimm should no longer be muted 14:22:43 Zakim, unmute me 14:22:43 Chimezie_Ogbuji should no longer be muted 14:22:48 subtopic: entailment 14:23:21 birte: lee's review missing, apart from that fine. had some comments from Enrico (Bolzano), will add some resp. editor's notes. 14:23:32 PROPOSED: Publish http://www.w3.org/2009/sparql/docs/entailment/ as WD 14:23:39 Lee: let's publish as is for now and discuss further upon published WD 14:23:43 +1 14:23:46 Zakim, mute me 14:23:46 bglimm should now be muted 14:23:49 +1 14:23:49 +1 14:23:52 +1 14:23:53 RESOLVED: Publish http://www.w3.org/2009/sparql/docs/entailment/ as WD 14:24:11 subtopic: rdf-http-update 14:24:27 I'm happy to review http update if not 14:25:01 Chime: don't remember if we specifically if we had reviews 14:25:21 Axel: just see that we didn't decide, because chime wasn't around. 14:25:34 chime: open comments are those from Tim Berners-Lee 14:26:18 Lee: steve, can you do a review on rdf-http-update? 14:26:31 Steve: I can see how far I get tonight 14:26:49 q+ 14:26:49 Lee: happy to publish contingent Steve's review 14:27:24 Andy: where are we on ?default ? should be noted in the doc. 14:27:41 Chime: it's noted, but haven't worked it in yet. 14:27:50 ISSUE: Handling of default graph in HTTP update protocol 14:27:50 Created ISSUE-63 - Handling of default graph in HTTP update protocol ; please complete additional details at http://www.w3.org/2009/sparql/track/issues/63/edit . 14:27:58 Lee: let's create a WG issue not to forget. 14:28:24 PROPOSED: Publish http://www.w3.org/2009/sparql/docs/http-rdf-update/ following Steve's OK or Steve's suggested changes and Chime's incorporation of them 14:29:00 +1 14:29:10 RESOLVED: Publish http://www.w3.org/2009/sparql/docs/http-rdf-update/ following Steve's OK or Steve's suggested changes and Chime's incorporation of them 14:29:13 +1 14:29:15 Leef: proposal means, if chime and steve can't agree, we'll just hold off publishing for now. 14:29:17 Zakim, mute me 14:29:17 Chimezie_Ogbuji should now be muted 14:29:28 subtopic: service description 14:29:46 http://www.w3.org/2009/sparql/docs/service-description-1.1/xmlspec.xml#sd-url 14:29:54 greg: updated status of the doc, other outstanding issue was formatting of schema information. 14:30:28 ... if people are ok with that http://www.w3.org/2009/sparql/docs/service-description-1.1/xmlspec.xml#sd-url formatting, I will propagate that throughout the doc. 14:31:13 ... I can commit that domain/range right after the call and then would be ready to go. 14:31:22 PROPOSED: Publish http://www.w3.org/2009/sparql/docs/service-description-1.1/xmlspec.xml as WD after pending formatting changes for schema info 14:31:33 +1 14:31:35 +1 14:31:35 +1 14:31:41 RESOLVED: Publish http://www.w3.org/2009/sparql/docs/service-description-1.1/xmlspec.xml as WD after pending formatting changes for schema info 14:32:40 leave Overview and test suite docs out for now. 14:32:46 -OlivierCorby 14:32:52 LeeF: leave Overview and test suite docs out for now. 14:32:53 topic: LET/BIND/assignment 14:33:05 http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0014.html 14:33:10 What is the date, we should use for the documents? 14:33:27 +OlivierCorby 14:34:35 OK 14:34:56 LeeF: we will follow up with editors on publication details with editors 14:35:03 straw poll: include LEt/BIND/whatever in SPARQL 1.1 14:35:07 ... on the date, etc. 14:35:25 +1 14:35:29 +1 (include) 14:35:29 0 14:35:33 -1 14:35:33 0 14:35:35 +1 14:35:35 -1 14:35:36 0 14:35:38 +1 (strong preferrence on *not* using LET as syntax) 14:35:47 0 14:35:49 q+ on syntax 14:35:49 0 14:35:54 ack me 14:37:09 Olivier: I think sparql is a graph match language, so Assignment is IMO not appropriate 14:37:30 Leef: we still have consensus in favor of inclusing it. 14:37:39 s/inclusing/including/ 14:37:49 two semantics questions 14:38:01 Leef: two issues on semantics 14:38:11 1) What happens if you try to assign/bind to a variable that is bound elsewhere in the query? 14:38:18 ... 1) assign/bind a variable that is bound elsewhere in the query 14:38:53 ... simple semantics vs. join semantics. 14:40:05 ... simple disallows that, other semantics creates a one column solution set and joins it with the current binding 14:40:10 q+ 14:40:58 ack AxelPolleres 14:40:58 AxelPolleres, you wanted to comment on syntax 14:41:31 extend is simple I believe 14:41:32 AxelPolleres: is the EXTEND operator simple semantics or join/filter semantics? 14:42:01 AndyS: it is EXTEND, but we have a syntactic restriction on the AS variable 14:42:04 ack SteveH__ 14:42:41 Axel: I am fine, under the syntactic restriction, if assign is just an extra syntactic sugar for Extend() 14:43:05 Steve: (??? can you please type in your question?) 14:43:44 LeeF: my idea for join semantics is a row-by-row join 14:43:59 Steve: is there a definition which does the join over the full table? 14:44:16 Andy: not sure whether that works, if it is a new variable 14:44:28 q? 14:44:37 Steve: that would be a separate piece of algebra. 14:44:44 very keen to avoid defining it in row-by-row terms 14:45:27 q+ 14:45:32 ack AxelPolleres 14:45:36 Ack - Not row-by-row : better is "add col to table" 14:46:00 -AlexPassant 14:46:59 I have a preference to only assign to new variables otherwise, it really exacerbates Olivier's concerns about taking a drastic step away from graph matching 14:47:21 +1 14:47:23 +??P2 14:47:24 Zakim, ??P2 is me 14:47:24 +AlexPassant; got it 14:47:27 Andy: if we want to make it order independent, then we need the filter effect. 14:48:01 ... let me try to write an example. 14:48:27 WHERE { ?x :p ?v . KW(?two := 2*?v) . ?x :p ?two } 14:48:50 (sorry about any emoticons) 14:48:51 Zakim, unmute me 14:48:51 Chimezie_Ogbuji should no longer be muted 14:49:13 Leef: we use this feature for UNIONs 14:49:14 WHERE { ?x ns:q ?v . KW(?two := 2*?v) . ?x ns:q: ?two } 14:49:30 WHERE { ?x ns:q ?v . ?x ns:q: ?two . KW(?two := 2*?v) . } 14:49:45 that would be illegal under simple semantics, right? 14:49:47 both 14:49:57 Andy: your example seems to be doable by normal FILTER as well? 14:50:22 s/Andy:/Andy,/ 14:51:08 q? 14:51:31 WHERE { ?x ns:q: ?two . KW(?two := 2) . } 14:51:42 WHERE { ?x ns:q: 2 } 14:52:13 WHERE { ?x ns:q: ?two . FILTER(?two = 2) 14:52:32 surely that would have to be sameTerm? 14:52:36 (yes) 14:53:11 straw poll: semantics? filter/join, simple, or 0 14:53:18 simple 14:53:23 strong pref for simple 14:53:28 0 14:53:30 simple 14:53:31 filter > simple > 0 14:53:33 simple 14:53:47 filter (mildly) 14:53:51 filter 14:53:56 0 / simple 14:54:42 LeeF: see a preference for the simple semantics, will require editors to look into what that means. 14:55:19 LET ( ?var := expr ) 14:55:24 BIND ( ?var := expr ) 14:55:34 LeeF: syntax... 14:55:35 BIND ( expr AS ?var ) 14:56:09 PLEASE LET ?var BIND TO expr 14:56:31 !@#$! NO! 14:56:32 The syntax restriction is not enforced by the algebra in (extend) 14:56:43 Por Favor | Please 14:57:43 -bglimm 14:57:55 it should behave like FILTER if it looks like FILTER, w.r.t. ordering 14:58:19 LeeF: semantics/order... do we want BIND to "execute" like a FILTER in the end of the group ort "in place"? 14:58:27 +bglimm 14:58:37 Zakim, mute me 14:58:37 bglimm should now be muted 14:58:45 Andy: in ARQ it is order-independent 14:58:52 WHERE { ?x ns:q ?v . KW(?two := 2*?v) . ?x ns:q: ?two } 14:58:59 in place semantics: the above is allowed 14:59:07 end of group semantics: the above is an error 14:59:20 it might be an error either way 14:59:24 ?two is still in scope 15:00:11 LeeF: we're out of time, let's fix this up first thing next week. 15:00:31 ... further suggestions on algebra for this appreciated on the mailing list! 15:00:41 ... thanks all, adjourned. 15:00:44 bye 15:00:45 bye 15:00:46 -Souri 15:00:46 -Chimezie_Ogbuji 15:00:47 -LeeF 15:00:48 -AlexPassant 15:00:49 -bglimm 15:00:49 -MattPerry 15:00:50 bye 15:00:51 -SteveH__ 15:00:53 -OlivierCorby 15:00:54 zakim, drop me 15:00:56 -Sandro 15:00:56 rrsagent, make records public 15:00:57 -AndyS 15:00:59 Ivan is being disconnected 15:01:01 -Ivan 15:01:05 -kasei 15:01:07 -AxelPolleres 15:01:32 SteveH_ has joined #sparql 15:07:42 and sinceok 15:07:44 oops 15:07:49 and sinceok yourself 15:07:54 i see 2 local.css files 15:07:56 one in /shared 15:07:58 one in shared/ 15:08:05 and one in service-description-1.1/ 15:08:05 and one in sha/red 15:08:14 there are local.css files in many of the document directories... 15:08:20 e.g. query-1.1/local.css 15:08:37 i suspect that the one in shared is a red herring 15:08:43 which is why I'm hesitant to have the shared xmlspec just blindly load it 15:09:12 it would be nice for the WG to have a single CSS file 15:09:22 entailment and property-paths seem to load their local.css explicitly 15:09:29 i wonder which of the local.css files are most recent 15:09:32 and query 15:10:00 i suspect that the browser will GET /path/query11/Overview.xml, then apply /path/shared/xmlspec.xslt which will tell it to load /path/query11/local.css 15:10:44 the documents aren't consistent on this. some load ../shared/local.css and some load just local.css. 15:10:55 ahh, interesting 15:12:16 merging local.css files seems ... unlikely to make me happy 15:14:16 hmm hmm 15:14:18 what a mess :) 15:15:06 so xmlspec.xsl in shared/ includes @import url("../shared/local.css"); 15:15:26 we could also have it @import url("local.css"); 15:15:30 but i don't know what that would break 15:17:02 maybe I should copy the other docs in their local importing of the (local) local.css? 15:17:06 (if that makes any sense :) 15:17:58 how do they do that? i see that entailment duplicates the xsl stuff, but I don't think that's a good idea 15:18:46 yeah, that's what I was looking at. 15:19:43 i think we just add import local.css after the other import 15:19:48 i'm going to do that and see what i break :) 15:20:55 OK, I checked that in 15:21:05 kasei, this should allow you to put whatever you want in the SD local.css and have it included 15:21:08 but i haven't tried it out :) 15:21:40 I think I just copied initially from query and I recently changed to use the local.css 15:22:10 I think we had some discussions before the last pub round and agreed that local should be local and include our own additions 15:22:13 my local.css file is full of junk I'm not using. I'm going to optimistically assume that removing it all will be OK. 15:22:49 sure, let's see 15:23:32